This week, the AI world is on fire! ๐ฅ OpenAI, Google, and Adobe have unveiled game-changing innovations that are set to redefine technology as we know it ๐โก. From powerful new AI tools and models to creative breakthroughs and smarter integrations, these giants are pushing the boundaries of what AI can do ๐๐จ. Stay ahead of the curve and discover how these developments will impact everything from business to daily life ๐ก๐. The AI revolution continues to accelerate โ donโt miss out!
00:00It's been a packed week full of cool updates in the AI world.
00:04Today, I'll tell you about the biggest and most interesting AI news from this week.
00:08We're looking at Google's hefty investment in AI firm Anthropic,
00:12OpenAI's new team tackling AI dangers,
00:16Google Maps' cool, immersive view feature,
00:18TikTok's AI-powered creative assistant for content creators,
00:22the S4IK Artist Award blending AI and art,
00:25Adobe's innovative project DubDubDub,
00:28an AI system discovering Supernova,
00:30the easy-to-use video crafter tool,
00:33UC's versatile mini-GPT5 model,
00:36and IBM's groundbreaking North Pole chip for AI efficiency.
00:39Okay, first, Google has recently decided to invest an additional $2 billion in the AI company Anthropic,
00:46which competes with OpenAI.
00:47This new investment is on top of the $550 million Google already invested in Anthropic earlier.
00:54The new funding is different from a previous deal where Google invested $500 million
00:58and promised to add another $1.5 billion later.
01:03Anthropic has also made a deal worth over $3 billion with Google Cloud a few months before this latest investment.
01:10Before Google's decision, Amazon had committed up to $4 billion to Anthropic,
01:14hoping to advance in the AI field against other tech companies.
01:17Amazon has already put in $1.25 billion, which could turn into a company share,
01:23and plans to possibly invest another $2.75 billion by early 2024.
01:29Anthropic was founded by Dario and Daniela Amodei, former engineers at OpenAI.
01:35They left OpenAI due to disagreements about safely developing AI.
01:39Previously, Anthropic received funding from Sam Bankman-Fried of FTX,
01:43but after FTX's downfall, it looked for new investors, including tech giants like Amazon and Google.
01:51Anthropic is working on an AI chatbot named Claude, similar to OpenAI's ChatGPT,
01:56and has gathered $7 billion this year to develop more advanced AI systems.
02:02All right, now OpenAI is forming a new team called Preparedness
02:06to tackle potential dangers of artificial intelligence.
02:09This team will look into risks from advanced AI like chemical, biological, and cybersecurity threats.
02:16They are concerned about how AI might be misused,
02:19and are working on strategies to ensure AI is safe and beneficial.
02:23OpenAI is aware that while AI has many advantages, it also comes with significant risks.
02:28They're now hiring people with various technical skills for this team,
02:32and have launched a challenge to encourage finding ways to prevent the misuse of AI.
02:36This initiative reflects growing concerns in the AI field about the technology's power
02:41and the need for stringent safety measures.
02:44Now, Google Maps has recently upgraded with a new Immersive View feature.
02:49This AI-powered tool transforms flat images into 3D views of landmarks, restaurants, and buildings,
02:55providing a more detailed perspective of various locations.
02:58It's now being integrated into navigation for 15 cities, including London, New York, and Tokyo,
03:04enhancing the experience for driving, walking, and cycling routes.
03:09Unlike traditional satellite or street view modes, Immersive View offers a 3D image of the entire route.
03:16It even allows users to see the route at different times or weather conditions using a slider powered by AI.
03:22Google Maps is also becoming more search-friendly, allowing general queries like artsy things nearby.
03:28This will display local attractions or specific user-uploaded photos tagged in locations.
03:33For using Immersive View, simply select it in the Google Maps app on a compatible city.
03:39Additionally, the Lens and Maps augmented reality feature, which shows nearby facilities like ATMs and cafes,
03:45is now available in 50 more cities.
03:48Okay, next.
03:49TikTok has introduced a new AI tool, the TikTok Creative Assistant,
03:54for brands and creators to use in the TikTok Creative Center with a TikTok for business account.
03:59This assistant offers smart guidance for creating TikTok content.
04:03It's loaded with TikTok-specific knowledge to provide relevant advice for creating ads or videos.
04:08The tool is versatile.
04:10It helps newcomers learn the basics of TikTok content creation,
04:14assists in researching successful TikTok ads,
04:16and aids in brainstorming and refining video scripts.
04:20The idea is to use artificial intelligence to continually inspire and assist in making effective TikTok content.
04:26You can start with suggested prompts from TikTok, but these are flexible.
04:31You control the final output, using, editing, or discarding the AI's suggestions as needed.
04:37All right, next.
04:38S4IK, a part of the multifaceted Azulic brand,
04:42is introducing the first S4IK Artist Award to promote artificial intelligence in art.
04:46This award offers $100,000 and a two-month creative residency at Azulic Omey in Mexico.
04:52Here, the winner can access advanced digital tools and local artisans for creating AI-based art.
04:58Runners-up also get art residencies.
05:01The award is a response to the growing intersection of AI and art.
05:04Applications are open until November 17, 2023.
05:08The award aligns with the opening of a new S4IK Museum project in Tulum,
05:13envisioned by Azulic's founder, Roth.
05:15This outdoor art space, part of the Azulic Basin,
05:18will open in 2024 and feature the winning artwork.
05:22Roth sees this initiative as a fusion of natural, ancestral, and artificial intelligence,
05:27marking a new era of global collaborative creativity.
05:30The S4IK Tulum and the Artist Award represent a commitment to pioneering in art,
05:36architecture, technology, and hospitality.
05:39Okay, now.
05:40Adobe revealed a new AI feature called Project Dub Dub Dub
05:44at their Adobe Max conference in October 2023, aimed at revolutionizing video dubbing.
05:50This feature, still in development, can dub videos automatically while keeping the original speaker's voice.
05:56This is part of Adobe's move to include more language AI tools in their products.
06:01Adobe's Zeyu Jin showcased this feature by dubbing voices into various languages
06:06and translating movie scenes without changing the original actor's voices.
06:10The technology, which is based on speech-to-speech translation,
06:14is still in its early stages, and it's not clear if Adobe is using its own AI or external models.
06:21The response to this new feature has been mixed,
06:23with some people concerned about the dubbing's clarity and timing.
06:27Despite this, many see its potential for improvement.
06:30This innovation is Adobe's way of helping brands reach global audiences
06:34by making videos more accessible across different languages.
06:38This approach is becoming increasingly common in the tech industry
06:42as companies seek new solutions for localization needs.
06:45Adobe is working alongside other big tech firms
06:48to integrate AI into creative and multilingual tools,
06:52although they note that Project Dub Dub Dub is not yet ready for release.
06:56Next, an international team led by Northwestern University developed a new AI system
07:01that automates finding, identifying, and classifying new supernova stars in the sky,
07:06replacing hours of human work.
07:08Previously, researchers spent over 2,200 hours examining supernova candidates.
07:14Now, this AI tool speeds up discoveries and reduces human error.
07:19The leader, Adam Miller, mentioned that for the first time,
07:22an automated system involving AI and robots detected a supernova
07:26and confirmed it with another telescope.
07:29His colleague, Nabeel Rehamtullah, emphasized this
07:32as the first fully automatic supernova classification.
07:36They tested this system, BTSBOT, on a new supernova, SN2023-TIK.
07:43The system independently handled the analysis, showing effectiveness in real-world conditions
07:48and allowing researchers to focus on other tasks.
07:51All right, next, new AI tool.
07:54Video Crafter is a free and open-source tool powered by diffusion models,
07:58a type of machine learning, enabling it to generate high-quality images and videos from text descriptions.
08:04It's perfect for those new to video editing, as it simplifies creating professional-looking videos.
08:10To use Video Crafter, you just write a description of the video, adjust settings like resolution, and hit generate.
08:17It creates a video from your words, which you can then edit and save in formats like MP4, MOV, and AVI.
08:25It's versatile, ideal for various video types, from educational to promotional or music videos.
08:30This user-friendly tool is expected to change how videos are made, offering a straightforward approach for anyone, regardless of their video editing skills.
08:40For installation and usage, their GitHub page provides all the necessary information.
08:45Video Crafter opens up new possibilities in video production, making it accessible and easy for everyone.
08:51Next, the University of California researchers have developed Mini-GPT-5, a new model that excels in both understanding and generating text and images.
09:01This model stands out because it uses generative Vulcans, a unique feature that helps it learn from images, a capability lacking in earlier text-focused models.
09:11Mini-GPT-5 combines these visual tokens with text, enabling it to create contextually relevant image captions.
09:19It's designed through a two-step process, first aligning text with quality visual features from large data sets, then ensuring coordination between text and visuals.
09:30This approach surpasses previous methods, balancing text image focus and optimizing training efficiency.
09:36Additionally, the model is enhanced with special training techniques to understand instructions better and excel in new tasks.
09:43The team's future work aims to expand its applications, overcoming the limitations of older models that treated image and text understanding separately, accessible and easy for everyone.
09:55And finally, IBM has unveiled North Pole, a new computer chip that significantly boosts the speed and efficiency of AI, particularly in image recognition.
10:05This chip is 22 times faster than current options.
10:08Traditional AI systems, like ChatGPT, often slow down due to their reliance on Internet data.
10:15But North Pole addresses this issue by using edge computing, which processes data closer to where it's needed.
10:21North Pole's design mimics the human brain, allowing its memory blocks and CPUs to interact quickly, ensuring rapid responses.
10:28While it outperforms competitors like NVIDIA GPUs in speed and processing, North Pole is not suited for training large AI models, such as ChatGPT.
10:39The research team is exploring ways to enhance its capabilities by linking multiple North Pole chips.
10:44Overall, North Pole represents a significant advancement in AI, promising faster and more efficient processing, especially in image recognition.
10:52Its current limitations notwithstanding, its potential for future improvements positions it as a game-changer in AI technology.
11:00Alright, that concludes this week's AI news.
11:03If you enjoyed this video, don't forget to hit the like button, subscribe to my channel, and click the bell icon to stay updated on all my future uploads.
11:11Thanks for tuning in, and I'll catch you in the next one.
11:22Thanks for tuning in, and I'll catch you in the next one.
